Backwards Walking for Knee Pain: Real Rehab or TikTok Gimmick?

Published October 19, 2026

Collage of a woman walking backwards along a flat park path, looking back over her shoulder to see where she is going, with the headline "Walk Backwards for Knees?"

Backwards walking is a reasonable add-on for knee osteoarthritis, but the research does not show it is a stand-alone fix, and one randomized trial could not show it was better than ordinary forward walking. The encouraging results come from small, short trials where it was added to physiotherapy, and the main practical risk is probably a fall rather than harm to your knee. If you try it, keep it slow, short and supported, and stop at any sharp or rising pain.

This article is general education, not medical advice. Knee pain has many causes, so have a physiotherapist or doctor check yours before you add backwards walking, especially if you have had a fall, have balance problems, or have had knee surgery.

Why backwards walking might be easier on the knee

Walking backwards (often called retro walking) changes how force travels through the leg. Lab studies back that up, though they measure loading in a single session, not whether anyone's pain improves.

Kneecap load. In a 1995 study of five male joggers (Flynn and Soutas-Little, Journal of Orthopaedic & Sports Physical Therapy), the peak compressive force behind the kneecap at self-selected speeds was about 3.0 body weights in backward running versus 5.6 in forward running, and it peaked later in the stance phase. That was running, in five men, so treat it as a clue and not a finding about walking.
Inner-knee load. In 2025, Zhang and colleagues (Bioengineering) put 32 people with medial knee osteoarthritis through one gait-lab session. Backwards walking cut the first peak of the knee adduction moment, a measure of sideways load on the inner side of the knee, by 33.3% compared with normal forward walking. It did not reduce the second peak, and the authors note that long-term effects are unknown.
Quadriceps activity. A 2019 study (Abdelraouf et al., Turkish Journal of Physical Medicine and Rehabilitation) recorded muscle activity in 20 young women with patellofemoral pain and 20 without. At 3 km/h, both the inner and outer quadriceps heads were more active walking backwards than forwards.

So there is a plausible mechanism: lower peak loads at some points in the step and busy thigh muscles. Mechanisms are not outcomes, though. The question for your knee is what happens to pain and function over weeks.

What the trials found

The best-known trial is Alghadir and colleagues (BMC Musculoskeletal Disorders, 2019): 68 adults aged 45 to 66 with mild to moderate knee osteoarthritis were split into retro walking, forward walking and control groups. Everyone received routine physiotherapy. The walking groups did 10 supervised minutes (with a 5-minute warm-up and cool-down) three days a week for six weeks at a comfortable speed, building toward 30 minutes if pain stayed below 3 out of 10.

Retro walking beat the physiotherapy-only group on all four outcomes measured. Pain fell by 1.8 points versus 1.0, WOMAC function scores improved by 4.8 versus 2.2, quadriceps strength rose by 1.7 kg versus 0.7 kg, and the timed up-and-go test improved by 0.6 seconds versus 0.1. But the study found no significant difference between retro and forward walking, and forward walking did not differ significantly from the physiotherapy-only group either, so the trial could not separate the two directions. The authors flag the small groups, the lack of long-term follow-up and the narrow age range.

A 2021 trial (Chen et al., Evidence-Based Complementary and Alternative Medicine) randomized 32 people aged 50 to 75 with knee osteoarthritis to four weeks of backwards walking plus usual care (needle-knife acupotomy, medication and routine exercise) or usual care alone. The backwards walkers reported less pain and better function, and the authors acknowledged the small sample.

Pooling trials gives the headline figure. As of October 2026, the most recent meta-analysis we found is Lin and colleagues (Physiotherapy Theory and Practice, published online in September 2025 and in print in January 2026). It combined 13 randomized trials of backwards walking added to conventional rehabilitation for knee osteoarthritis. The pooled effect sizes were close to 1.0 for both pain (Hedges' g of about -1.0, meaning group averages roughly one standard deviation apart) and disability, which the authors described as moderate to large. Effects were bigger in trials that used a comfortable walking speed and a treadmill. The authors said future work should test backwards walking on its own, look at long-term results and assess the certainty of the evidence.

How much to trust that evidence

Take the results seriously but not at face value.

It is an add-on, not a replacement. The pooled trials tested backwards walking added to a rehabilitation program, not backwards walking by itself. Extra exercise time may help on its own, which is why the forward-walking arm in Alghadir's trial matters: it did not differ significantly from the backwards arm, although that trial was too small to settle the question.
The trials are short and small. The two trials above lasted four and six weeks, had 32 and 68 participants, and had no long-term follow-up. Pain is self-reported, and people know which exercise they are doing.
The people studied were fairly narrow. Alghadir's participants had mild to moderate osteoarthritis (radiographic grades 1 to 3) and were aged 45 to 66. Between them, the two trials excluded people with previous knee surgery, a leg fracture or acute fluid in the knee joint. The trials above do not tell you much about tendon problems or post-surgery knees, and for kneecap pain the support in this article is lab-based only.
Exercise as a whole is the best-supported treatment. A 2024 Cochrane review of 139 randomized trials with 12,468 people (Lawford et al.) compared exercise with no treatment, usual care or limited education. It found that exercise may reduce knee osteoarthritis pain (about 13 points on a 0 to 100 scale, low-certainty evidence) and likely improves function (about 12.5 points, moderate-certainty). The authors judged these benefits to be of uncertain clinical importance, and most trials were at unclear or high risk of bias. The 2019 American College of Rheumatology and Arthritis Foundation guideline (Kolasinski et al., published 2020) gives exercise a strong recommendation.

The fair reading is that walking backwards is one of many exercises that may help a knee with osteoarthritis, and the unproven claim is that the backwards direction adds something special. If you already strengthen your legs and walk, backwards walking is an optional extra. See our guide to knee pain when squatting for the lifting side, and why training itself protects joints.

Falling is the risk to plan for

You cannot see where you are going, and your balance reactions are less practiced backwards. A 2025 case-control study of 483 community-dwelling older adults (Polo-Ferrero et al., Journal of Functional Morphology and Kinesiology) found that people who had fallen in the previous 12 months took longer on a 3-meter backward walk (6.8 versus 5.1 seconds on average). It was observational, and on its own the test was a poor predictor of who had fallen (an area under the curve of 0.655, which the authors' own scale rates as poor). So it suggests that backwards walking is harder for people with poor balance, not that it causes falls.

Both knee trials described above reported no serious adverse events. But the walking was supervised or checked, at a comfortable pace, in people who met entry rules. Alghadir's sessions were supervised, while in Chen's trial only the first session was, with the rest done at home and telephone check-ups. Chen's trial also required people to be able to stand for 30 seconds without a support device. Your own session will probably be less controlled than that, for example on a gym treadmill at the end of a busy evening.

Backwards walking can also be harder on the cardiovascular system than it looks. In a 2004 study of 27 young adults walking at 2.5 mph on inclines of 5% to 10% (Hooper et al., Journal of Orthopaedic & Sports Physical Therapy), percent of maximum heart rate and percent of maximum oxygen uptake were 17% to 20% higher backwards than forwards. So a pace that feels easy forwards may take noticeably more effort in reverse, at least on an incline.

A cautious way to start

If your clinician is happy, build up in a way that follows the trial approach: comfortable speed, pain under 3 out of 10, time added slowly. The schedule below is an adaptation, not a protocol from any trial.

Weeks 1 and 2. Two or three days a week, add 1 to 2 minute backwards segments inside a normal walk, for about 5 minutes in total.
Weeks 3 and 4. Three days a week, build to segments of 3 to 5 minutes and about 10 minutes in total.
Week 5 onward. One 10-minute block, with a 5-minute warm-up and cool-down as in the trials. Add time only if pain stayed under 3 out of 10 during and after the session and the knee is no worse the next day. The trials allowed up to 30 minutes.

For a worked example, suppose you walk 20 minutes on a treadmill three times a week. In week 1, spend minutes 5 to 6, 10 to 11 and 15 to 16 going backwards at the lowest belt speed with both hands on the rails, and stop the belt fully before you turn around. That is about 6 minutes of backwards walking in the session.

Treadmill. Clip the safety key to your clothes, hold the handrails, drop the speed well below your forward pace, raise it in the smallest steps the machine allows, and keep the incline flat. Do not copy the steep incline of trends such as the 12-3-30 workout in reverse. Forward treadmill walking is a sensible way to warm up first.
Floor. Use a flat, clear stretch such as a quiet running-track lane or empty hallway, with a helper who watches the way you are heading if you can. Avoid crowds, curbs and uneven ground, and check over your shoulder before each pass.

Who should check with a clinician first

Get advice before trying this, and do not do it unsupervised, if any of these apply:

Knee pain after an injury, twist or fall, or pain that has not improved within a few weeks, which the NHS says to take to a GP. UK NHS guidance says to get urgent advice (from 111) if your knee is very painful, you cannot move it or put any weight on it, it is badly swollen or has changed shape, it locks, gives way or painfully clicks, or you have a very high temperature with redness or heat around the knee. Those signs need assessment, not exercise.
Balance or falls problems, dizziness, numb feet, poor vision, or a history of falls. A clinician may suggest supervised sessions first, or a different exercise.
Recent knee or hip surgery, or a fracture in the leg. The trials above excluded people with previous knee surgery or a leg fracture, so there is little evidence for these groups. Surgeons and physiotherapists set the timing.
Heart or lung conditions, since backwards walking can raise the effort at a given speed.

A clinician is the person to decide what is actually wrong (osteoarthritis, a meniscal problem, kneecap pain or something else), whether the exercise suits that diagnosis, whether you can do it safely alone, and how fast to progress. Stop the session if pain climbs above 3 out of 10, and get checked if the pain does not settle, swelling follows, or the knee catches or gives way.

Common mistakes, and where it fits in your week

Going too fast. The pooled analysis found bigger effects at comfortable speed, and a fall costs more than a slow session.
Jumping to 30 minutes. The trials progressed only while pain stayed low.
Judging after three sessions. The trials ran four to six weeks.
Treating a locking or giving-way knee as something to walk off. Mechanical symptoms need a diagnosis.
Dropping strength work. The physiotherapy that every group in Alghadir's trial received included quadriceps strengthening such as straight leg raises, terminal knee extension, semi-squats and the leg press.

For most lifters, backwards walking fits as 5 to 10 minutes at the end of a session or on a lighter day, built on top of a leg program you can do with little pain. It does not replace walking itself either; our comparison of 10,000 steps versus weight training covers why daily steps still matter, and training after 50 covers the strength side for older lifters.

Easy Reps is a free workout tracker for Apple devices with 900+ exercises, including treadmill walking, and it supports timed sets. If you add a backwards block to a routine, the workout history and calendar let you look back at your past sessions, which helps when a progression asks you to add time slowly. It does not record pain scores, so write the 0 to 10 number for each session somewhere else, such as your phone's notes.

Bottom line

Backwards walking has a plausible mechanism and promising trial results for mild to moderate knee osteoarthritis, but the trials are small, short and mostly test it as an add-on to rehabilitation, and one could not show it was better than forward walking. Exercise in general is the best-supported treatment, so keep your strength work and regular walking, and treat walking backwards as an optional extra.

If you try it, go slow, hold the rails or use a helper, build time gradually while pain stays under 3 out of 10, and see a clinician first if your knee is injured, locks, swells or gives way, or if balance is a concern.